ISTIO_CNI_ARGS="--set cni.cniBinDir=/home/kubernetes/bin"
fi
iop kube-system istio-cni $IBASE/istio-cni/ ${ISTIO_CNI_ARGS}
```

TODO. It is possible to add Istio-CNI later, and gradually migrate.

### Install Control plane

This can run in any cluster. A mesh should have at least one cluster should run Pilot or equivalent XDS server,

* All traffic *ingressing* a pod on port 15008 in the mesh is assumed to be HBONE, and should be redirected to the node-local ztunnel on port 15008 (more on this later).
* All other traffic *ingressing* a pod in the mesh should be redirected to the node-local ztunnel on port 15006, regardless of intended original destination port.

  // The check can be enabled immediately when deploying a driver.
  // In that case, provisioning new volumes with late binding
  // will pause until the driver deployment has published
  // some suitable CSIStorageCapacity object.
  //
  // Alternatively, the driver can be deployed with the field
  // unset or false and it can be flipped later when storage
  // capacity information has been published.
  //

  // The check can be enabled immediately when deploying a driver.
  // In that case, provisioning new volumes with late binding
  // will pause until the driver deployment has published
  // some suitable CSIStorageCapacity object.
  //
  // Alternatively, the driver can be deployed with the field
  // unset or false and it can be flipped later when storage
  // capacity information has been published.
  //

  // Errors may occur at different points while syncing config. Earlier errors (e.g. download or
  // checkpointing errors) will not result in a rollback to LastKnownGood, and may resolve across
  // Kubelet retries. Later errors (e.g. loading or validating a checkpointed config) will result in
  // a rollback to LastKnownGood. In the latter case, it is usually possible to resolve the error
  // by fixing the config assigned in Spec.ConfigSource.

to generate a new secret and push the updated certificate to Envoy. This ensures that we do not permanently watch certificates even after
Envoy has stopped requested them; if there are no subscriptions they update will be ignored. If Envoy later watches these certificates again,
a new one will be generated on demand.

## Configuration

| Variable | Description |
| - | - |
|CA_ADDR|Address of CA, defaults to discoveryAddress|

#### Endpoints

Endpoints have an optimized code path, as they are by far the most frequently updated resource - in a steady cluster, this will often be the *only* change, caused by scale up/down.

  // in which the component responsible for the first finalizer in the list is
  // waiting for a signal (field value, external system, or other) produced by a
  // component responsible for a finalizer later in the list, resulting in a deadlock.
  // Without enforced ordering finalizers are free to order amongst themselves and
  // are not vulnerable to ordering changes in the list.
  // +optional
  // +patchStrategy=merge